Stuff I Use

My current setup
I work on an LG UltraFine Display with my Apple MacBook Pro 14" M4 Max, using a Logitech MX Master 2S mouse on a self-assembled Jin Office Electric Standing Desk. My chair is the Featherlite Liberate, lit by an Ikea TΓ GARP lamp.
EDC
Headphones
Sony WH-1000XM3
Earphones
AirPods 3, CMF Buds 2 Pro
Speaker
Marshall Willen II
Camera
Sony A6400
Primary Lens
Sigma 30mm f/1.4
Others
iPad, Kindle, Apple Watch
Softwares
For browsing, I use Arc β after a long detour through every alternative worth trying, I'm back, sticking with it until it's discontinued for good, though still on the lookout for a WebKit-based alternative like Nook. My email default is Apple Mail, though I've been trying Extra β the mail app from the folks behind Pinterest β and loving it so far. Raycast has been my preferred launcher since 2020, and I love it. It's a staple, with over 50 custom settings tuned in over the years. Not upgrading to v2 until they force it.
I take notes in Bear and handle to-dos with Apple Reminders β I switched over from TeuxDeux not too long ago. Notion Calendar keeps my schedule for now. Notion has already shut down Mail, though, and will likely sunset Calendar too β so I may end up back on Apple Calendar soon. For AI assistants, it's Claude day-to-day β I'm out of both the ChatGPT and Perplexity ecosystems now β with Gemini for Workspace and collaboration.
CoTabby's the open-source alternative to Cotypist. I'm always dictating over typing these days, so TypeWhisper does most of the work, with s1-mini handling post-processing through a small extension I built myself. I keep coming back to TypeWhisper not just because it's open-source, but for its workflow capabilities, extensions, and an active open-source scene that ships updates and fixes more often than the rest of the OSS dictation crowd. I still have a soft spot for how simple Hex was (retired now), and Handy is solid too. What I won't do is pay a subscription for dictation apps like Superwhisper or Whispr Flow β they're not local, the privacy trade-off isn't worth it, and cloud dictation eats far too much of your machine's resources for what it does. If you need online inference, run Whisper on Groq instead. That said, credit where it's due β Superwhisper putting out s1-mini as an open-source model has been a big help.